Mathematical Models for Predicting Solar Irradiance at a Site for Milk Solar Refrigeration Systems

Abstract

The mismatch between solar energy availability and the cooling load energy demands for AC solar refrigeration systems in different geographical locations complicates the design and sizing of milk solar refrigeration systems components. This is caused by variation seasonal solar insolation and different levels of global solar insolation. In this study, three different sizes AC milk solar refrigeration systems, have been investigated for maximum cooling loads developed from the refrigeration systems when exposed to varying levels of solar insolation in Nakuru Kenya. Regression models were developed for predicting maximum cooling loads delivered from the milk solar refrigeration systems based on available mean daily solar insolation of the location. The predictive models developed forms as useful tools in the design and sizing of milk solar refrigeration components based on solar insolation available at any global location. Three Solar refrigeration systems were fitted with AC reciprocating compressors of capacities’; 350 W, 250 W, 200 W and were investigated for maximum cooling loads under varying mean daily solar radiations. Four PV panels each of 200 Wp connected via an inverter provided the power required to operate the compressors in each of the refrigeration system. An innovative control unit operated the refrigeration systems dependent on the solar insolation level available in the day. Temperature profiles of water placed in the central water can, and the amount of ice formed were used to determine the maximum cooling load of each refrigeration system with, based on solar radiation available. The regression cooling curve generated by each system was used in developing the mathematical cooling load prediction models based on available solar insolation of Nakuru. The results showed that the maximum cooling loads obtained from the solar refrigeration systems dependent on the annual mean daily solar insolation of a specific location and the capacity of the refrigeration system compressor. The mathematical models showed a strong correction of coefficient of between 0.958 and 0.908when validated with actual solar refrigeration cooling loads.